Memory Defect Management via Block Segmentation and Density Mode Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional memory management techniques often lead to premature retirement of entire blocks in memory systems due to failures, which can result in underutilization of functional memory cells, especially when distinguishing between early-life and late-life failures is not adequately addressed.
Innovation Solution
Implementing a defect management system that includes failure detection logic to differentiate between extrinsic and intrinsic failures, allowing for partial block reconfiguration, such as splitting bands into sub-blocks or converting storage density modes, to extend the life of memory systems without retiring entire blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If entire blocks are retired due to failures, then reliability is improved, but memory utilization deteriorates
Solution Approach 1:
The patent segments a memory block into multiple sub-blocks when a failure is detected. Instead of retiring the entire block, the system identifies the specific failed sub-block and isolates it, allowing the remaining functional sub-blocks to continue operating. This segmentation approach resolves the contradiction by maintaining reliability (through failure isolation) while preserving memory utilization (by keeping healthy portions active).
Solution Approach 2:
The patent extracts or removes only the defective portion (failed sub-block) from the memory block while leaving the rest intact. The failed sub-block is retired individually, and the remaining functional sub-blocks continue to serve data storage needs. This extraction principle resolves the contradiction by eliminating the harmful effect of the failure without unnecessarily retiring the entire block, thus maintaining both reliability and memory utilization.
2Reliability
If blocks are retired early to ensure reliability, then failure risk is reduced, but memory capacity is lost
Solution Approach 1:
By segmenting blocks into sub-blocks, the patent enables granular failure management. When a failure occurs, only the affected sub-block is retired rather than the entire block. This prevents premature retirement of healthy memory capacity while maintaining reliability through systematic failure isolation and sub-block-level management.
Solution Approach 2:
The patent changes the operational parameters of memory blocks by dynamically adjusting which sub-blocks are active based on failure detection. The system modifies the operational state from whole-block retirement to sub-block-level retirement, optimizing the balance between reliability and capacity utilization by retiring only the minimum necessary portions.
3Stability of the object's composition
If whole blocks are retired for failures, then system stability is improved, but memory efficiency deteriorates
Solution Approach 1:
The patent implements segmentation of memory blocks into sub-blocks, enabling stable failure management at a granular level. When failures occur, the system maintains stability by isolating and retiring only the affected sub-blocks while keeping the rest of the block operational. This approach preserves memory efficiency by preventing unnecessary retirement of healthy portions, thus resolving the contradiction between stability and efficiency.
Solution Approach 2:
The patent introduces dynamic sub-block retirement mechanisms that adapt to actual failure conditions. The system dynamically determines which sub-blocks to retire based on real-time failure detection, rather than applying static whole-block retirement policies. This dynamic approach maintains system stability through systematic failure management while optimizing memory efficiency by preserving functional capacity.
Data Source
AI summary
Defect management logic extends a useful life of a memory system. For example, as discussed herein, failure detection logic detects occurrence of a failure in a memory system. Defect management logic determines a type of the failure such as whether the failure is an infant mortality type failure or a late-life type of failure. Depending on the type of failure, the defect management logic performs different operations to extend the useful life of the memory system. For example, for early life failures, the defect management logic can retire a portion of the block including the failure. For late life failures, due to excessive reads/writes, the defect management logic can convert the failing block from operating in a first bit-per-cell storage density mode to operating in a second bit-per-cell storage density mode.